Twelve subjects who met diagnostic criteria for episodic tension-type headache and nine subjects who rarely or never suffered from headaches wore a computer-controlled electromyographic (EMG) activity recorder in their natural environment for 48 to 96 consecutive hours. EMG activity of the posterior neck or frontal muscles was recorded 24 hr per day. During waking hours, subjects rated their perceived levels of stress, pain, and negative affect at 30-min intervals. The EMG activity of headache and control subjects did not differ significantly, and EMG activity did not covary with stress, pain, or negative affect. Cross-correlations among EMG activity, pain, and stress revealed little evidence of leading, contemporaneous, or lagging relationships. Interrupted time series analysis showed no consistent muscle hyperactivity during a headache attack compared to a headache-free baseline period.